From: "Wu, Fei" <fei2.wu@intel.com>
To: Richard Henderson <richard.henderson@linaro.org>,
<qemu-devel@nongnu.org>, <alex.bennee@linaro.org>
Cc: "Vanderson M. do Rosario" <vandersonmr2@gmail.com>,
Paolo Bonzini <pbonzini@redhat.com>
Subject: Re: [PATCH v12 02/15] accel: collecting TB execution count
Date: Thu, 25 May 2023 08:45:07 +0800 [thread overview]
Message-ID: <cc040c82-3c2d-f781-1168-4cb76d90d2e9@intel.com> (raw)
In-Reply-To: <fd8a711d-7940-c990-bbed-7c4d9a46a664@linaro.org>
On 5/25/2023 1:02 AM, Richard Henderson wrote:
> On 5/24/23 06:35, Wu, Fei wrote:
>> On 5/23/2023 8:45 AM, Richard Henderson wrote:
>>> On 5/18/23 06:57, Fei Wu wrote:
>>>> +void HELPER(inc_exec_freq)(void *ptr)
>>>> +{
>>>> + TBStatistics *stats = (TBStatistics *) ptr;
>>>> + tcg_debug_assert(stats);
>>>> + ++stats->executions.normal;
>>>> +}
>>> ...
>>>> +static inline void gen_tb_exec_count(TranslationBlock *tb)
>>>> +{
>>>> + if (tb_stats_enabled(tb, TB_EXEC_STATS)) {
>>>> + TCGv_ptr ptr = tcg_temp_new_ptr();
>>>> + tcg_gen_movi_ptr(ptr, (intptr_t)tb->tb_stats);
>>>> + gen_helper_inc_exec_freq(ptr);
>>>> + }
>>>> +}
>>>
>>> This is 3 host instructions, easily expanded inline:
>>>
>>> --- a/accel/tcg/translator.c
>>> +++ b/accel/tcg/translator.c
>>> @@ -11,6 +11,7 @@
>>> #include "qemu/error-report.h"
>>> #include "tcg/tcg.h"
>>> #include "tcg/tcg-op.h"
>>> +#include "tcg/tcg-temp-internal.h"
>>> #include "exec/exec-all.h"
>>> #include "exec/gen-icount.h"
>>> #include "exec/log.h"
>>> @@ -18,6 +19,30 @@
>>> #include "exec/plugin-gen.h"
>>> #include "exec/replay-core.h"
>>>
>>> +
>>> +static void gen_tb_exec_count(TranslationBlock *tb)
>>> +{
>>> + if (tb_stats_enabled(tb, TB_EXEC_STATS)) {
>>> + TCGv_ptr ptr = tcg_temp_ebb_new_ptr();
>>> +
>>> + tcg_gen_movi_ptr(ptr,
>>> (intptr_t)&tb->tb_stats->executions.normal);
>>> + if (sizeof(tb->tb_stats->executions.normal) == 4) {
>>> + TCGv_i32 t = tcg_temp_ebb_new_i32();
>>> + tcg_gen_ld_i32(t, ptr, 0);
>>> + tcg_gen_addi_i32(t, t, 1);
>>> + tcg_gen_st_i32(t, ptr, 0);
>>> + tcg_temp_free_i32(t);
>>> + } else {
>>> + TCGv_i64 t = tcg_temp_ebb_new_i64();
>>> + tcg_gen_ld_i64(t, ptr, 0);
>>> + tcg_gen_addi_i64(t, t, 1);
>>> + tcg_gen_st_i64(t, ptr, 0);
>>> + tcg_temp_free_i64(t);
>>> + }
>>> + tcg_temp_free_ptr(ptr);
>>> + }
>>> +}
>>> +
>>> bool translator_use_goto_tb(DisasContextBase *db, target_ulong dest)
>>> {
>>> /* Suppress goto_tb if requested. */
>>>
>>>
>>> I'm not expecially keen on embedding the TBStatistics pointer directly
>>> like this; for most hosts we will have to put this constant into the
>>> constant pool. Whereas the pointer already exists at tb->tb_stats, and
>>> tb is at a constant displacement prior to the code, so we already have
>>> mechanisms for generating pc-relative addresses.
>>>
>>> However, that's premature optimization. Let's get it working first.
>>>
>> Richard, have you reviewed the whole series? I will integrate your
>> change to next version.
>
> No, it's difficult to see what's going on.
> In your next revision, please remove CONFIG_PROFILER entirely first,
> which was what I was planning to do locally.
>
OK, I will update it.
Thanks,
Fei.
>
> r~
next prev parent reply other threads:[~2023-05-25 0:46 UTC|newest]
Thread overview: 28+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-05-18 13:57 [PATCH v12 00/15] TCG code quality tracking Fei Wu
2023-05-18 13:57 ` [PATCH v12 01/15] accel/tcg: introduce TBStatistics structure Fei Wu
2023-05-18 14:16 ` Wu, Fei
2023-05-18 13:57 ` [PATCH v12 02/15] accel: collecting TB execution count Fei Wu
2023-05-23 0:45 ` Richard Henderson
2023-05-23 1:48 ` Wu, Fei
2023-05-23 13:08 ` Wu, Fei
2023-05-24 13:35 ` Wu, Fei
2023-05-24 17:02 ` Richard Henderson
2023-05-25 0:45 ` Wu, Fei [this message]
2023-05-25 13:27 ` Wu, Fei
2023-05-18 13:57 ` [PATCH v12 03/15] accel: collecting JIT statistics Fei Wu
2023-05-18 13:57 ` [PATCH v12 04/15] accel: replacing part of CONFIG_PROFILER with TBStats Fei Wu
2023-05-18 13:57 ` [PATCH v12 05/15] accel/tcg: move profiler dev_time to tb_stats Fei Wu
2023-05-18 13:57 ` [PATCH v12 06/15] accel/tcg: convert profiling of restore operations to TBStats Fei Wu
2023-05-18 13:57 ` [PATCH v12 07/15] accel/tcg: convert profiling of code generation " Fei Wu
2023-05-18 13:57 ` [PATCH v12 08/15] accel: adding TB_JIT_TIME and full replacing CONFIG_PROFILER Fei Wu
2023-05-18 13:57 ` [PATCH v12 09/15] debug: add -d tb_stats to control TBStatistics collection: Fei Wu
2023-05-18 13:57 ` [PATCH v12 10/15] monitor: adding tb_stats hmp command Fei Wu
2023-05-18 14:11 ` Wu, Fei
2023-05-18 13:57 ` [PATCH v12 11/15] tb-stats: reset the tracked TBs on a tb_flush Fei Wu
2023-05-18 13:57 ` [PATCH v12 12/15] Adding info [tb-list|tb] commands to HMP (WIP) Fei Wu
2023-05-18 14:30 ` Wu, Fei
2023-05-18 13:57 ` [PATCH v12 13/15] tb-stats: dump hot TBs at the end of the execution Fei Wu
2023-05-18 13:57 ` [PATCH v12 14/15] configure: remove the final bits of --profiler support Fei Wu
2023-05-18 13:57 ` [PATCH v12 15/15] docs/tb-stats: add how to Fei Wu
2023-05-22 10:25 ` Thomas Huth
2023-05-22 12:51 ` Wu, Fei
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=cc040c82-3c2d-f781-1168-4cb76d90d2e9@intel.com \
--to=fei2.wu@intel.com \
--cc=alex.bennee@linaro.org \
--cc=pbonzini@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=richard.henderson@linaro.org \
--cc=vandersonmr2@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).